Staff Software Engineer (Internal Tools)

Join Tubi, a leading ad-supported streaming service, as a Staff Software Engineer. In this role, you will drive the technical evolution of our AI-centric business platforms, focusing on ad revenue operations. You will confront architectural challenges, define and implement architectural patterns, and validate their effectiveness. You will also own the platform's most complex subsystems, establish clear system boundaries, and ship consistently. Additionally, you will mature the platform's AI systems and raise the engineering bar through code review and design discussions.

• Conduire l'évolution technique des plateformes commerciales centrées sur l'IA, en intégrant l'IA non déterministe dans les flux de travail.
• Définir et piloter l'architecture technique de la plateforme d'IA, en veillant à ce que les systèmes soient modulaires, testables et évolutifs.
• Maturer les systèmes d'IA de la plateforme, en assurant une validation systématique, des boucles de rétroaction et une gestion des coûts.
• Good technical judgment under uncertainty. Pragmatic trade-offs, knowing when "good enough" is right, and not over-engineering when the problem doesn't call for it
• Deep expertise in distributed systems design, fault tolerance, and building reliable systems from unreliable components
• Demonstrated bias toward shipping. Track record of turning ambiguous problems into working systems on a reasonable timeline — not just designs or proposals
• 8+ years of software engineering experience, with demonstrated ability to own and deliver complex systems end-to-end — from design through production operation
• Strong full-stack engineering fundamentals. Our stack: Node.js/TypeScript (backend, Express), React/TypeScript (frontend), PostgreSQL with Drizzle ORM, Python for ML pipeline components, OpenAI and Google Gemini APIs
• Tubi expects engineers at all levels to leverage AI tools (Claude Code, Codex, Cursor, MCP integrations) to accelerate delivery, testing, and documentation
• Familiarity with LLM application patterns: prompt engineering, multi-model orchestration, agentic tool use, and cost optimization for high-volume LLM workloads
• Experience building over heterogeneous enterprise data sources with varying freshness, schemas, and access patterns
• Experience with document AI or structured data extraction from unstructured sources (documents, image, natural language) at scale
• Experience designing systems that handle non-deterministic or probabilistic outputs (ML models, LLM pipelines, or similar) with appropriate confidence scoring and fallback strategies
• Background in ad tech, revenue operations, or media operations — order management, pricing logic, deal lifecycle
• For this role specifically, AI goes further — it is not a feature on this platform but the reason the platform exists. You don't need to arrive as an AI expert, but you need to develop deep fluency quickly and apply it with production-grade discipline:
• Trust Engineering: Building systems where human operators can progressively delegate to AI — and where the blast radius is contained when AI is wrong
• Cost Discipline: Treating AI spend as an engineering problem — optimizing for accuracy per dollar, not just accuracy alone
• Pragmatic Adoption: Evaluating new capabilities with a builder's eye — what to use now, what's hype, what to wait for
• AI as Leverage: Thinking about AI as a force multiplier for a small team. The systems you build should measurably expand what the team can accomplish